Business risk prediction method and device for setting and adjusting risk threshold value based on profit and loss balance point

Abstract

The invention relates to the technical field of artificial intelligence, discloses a business risk prediction method and device for setting and adjusting a risk threshold value based on a profit and loss balance point, equipment, a medium and a product, and aims to solve the problems that the balance of cost and profit is not considered, dynamic adjustment capability is lacked, the risk threshold value cannot be dynamically adjusted, and the risk threshold value cannot be dynamically adjusted in the prior art. And the accuracy and flexibility of credit risk control are poor. The method comprises the following steps: constructing an initial prediction model, and collecting and training the initial prediction model based on historical business data to obtain a business risk prediction model; calculating a profit and loss balance point of the business based on the historical business data, and setting an adjustment risk threshold based on the profit and loss balance point; and obtaining a to-be-predicted target service, calling the service risk prediction model to perform service information prediction on the target service, obtaining a service information prediction value, comparing the service information prediction value with the set and adjusted risk threshold, and determining a final service risk result. According to the method, the accuracy and flexibility of the business risk prediction method are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and in particular to a business risk prediction method and device based on setting and adjusting a risk threshold value at a break-even point, equipment, medium and product. BACKGROUND

[0002] In the fields of public safety, transportation, engineering, finance, etc., risk control is a crucial and indispensable core link. In the financial field, in credit business, etc., risk control constitutes the core guarantee for the stable operation of financial institutions. Among them, the setting of the risk threshold value in the risk control model for credit has a key influence on the accuracy of credit decision-making, the effectiveness of risk control, and the income level of financial institutions.

[0003] However, the traditional credit risk threshold setting method has many limitations and cannot meet the needs of modern financial institutions for fine risk management. For example, in existing technologies, financial institutions usually set a fixed risk threshold value based on industry experience or historical data. When predicting business risks, the prediction results are compared with the fixed risk threshold value to determine whether there is a risk. The disadvantage of this scheme is that it cannot be dynamically set or adjusted according to changes in market environment, and when business information fluctuates, the balance between cost and benefit is not considered. This rigid setting cannot adapt to market changes in a timely manner, lacks dynamic adjustment capability, is prone to cause deviation in risk assessment, and thus reduces the accuracy of risk prediction and affects the prediction effect. Therefore, there is an urgent need for a business risk prediction method that can dynamically adjust the risk threshold value to improve the accuracy and decision flexibility of credit risk control. SUMMARY

[0004] The main purpose of the present application is to solve the technical problem that the existing business risk prediction method does not consider the balance between cost and benefit, lacks dynamic adjustment capability, cannot dynamically adjust the risk threshold value, and results in poor accuracy and flexibility of business risk control.

[0005] The first aspect of the present application provides a business risk prediction method based on setting and adjusting a risk threshold value at a break-even point, comprising: constructing an initial prediction model, collecting and training the initial prediction model based on historical business data to obtain a business risk prediction model; calculating the break-even point of the business based on historical business data, and setting and adjusting the risk threshold value based on the break-even point; obtaining a target business to be predicted, calling the business risk prediction model to predict business information of the target business, and obtaining a business information prediction value; comparing the business information prediction value with the set and adjusted risk threshold value to determine the final business risk result.

[0006] Optionally, in the first implementation manner of the first aspect, the calculating the break-even point of the business based on the historical business data comprises: obtaining historical business data of a first business period, and calculating a first break-even point of the historical business data of the first business period; determining a second business period based on a risk prediction requirement, and calculating a second break-even point of the second business period based on the first break-even point; wherein the second business period has a time length less than the first business period.

[0007] Optionally, in the second implementation manner of the first aspect, the calculating the second break-even point of the second business period based on the first break-even point comprises: obtaining a risk critical point of a business risk prediction model corresponding to the break-even point of the first business period based on the historical business data, to obtain a critical score range; searching for a second break-even point corresponding to the second business period based on the critical score range; the setting the adjustment risk threshold based on the break-even point comprises: setting a risk threshold corresponding to the second business period based on the second break-even point.

[0008] Optionally, in the third implementation manner of the first aspect, the first business period is 12 periods or 24 periods, and the second business period is 6 periods or 3 periods.

[0009] Optionally, in the fourth implementation manner of the first aspect, after the searching for the second break-even point corresponding to the second business period based on the critical score range, the method further comprises: judging whether the target business content has a definite business return period; if not, obtaining a historical loan amount proportion in a preset period, and performing rate dimension weighting on the second break-even point based on the historical loan amount proportion as a weight.

[0010] Optionally, in the fifth implementation manner of the first aspect, before the comparing the business information prediction value and the risk threshold set after the adjustment to determine a final business risk result, the method further comprises: obtaining a business risk preference of a target business to be predicted, and selecting a risk strategy based on the business risk preference; when the risk strategy is a conservative strategy, lowering the risk threshold; when the risk strategy is a fund strategy, raising the risk threshold.

[0011] The second aspect of the present application provides a business risk prediction device based on setting and adjusting a risk threshold value based on a break-even point, comprising: a model construction module, configured to construct an initial prediction model, collect and train the initial prediction model based on historical business data to obtain a business risk prediction model; a threshold adjustment module, configured to calculate a break-even point of a business based on historical business data, and set and adjust a risk threshold value based on the break-even point; a prediction execution module, configured to obtain a target business to be predicted, call the business risk prediction model to perform business information prediction on the target business, and obtain a business information prediction value; a result output module, configured to compare the business information prediction value with the set and adjusted risk threshold value, and determine a final business risk result.

[0012] The third aspect of the present application provides a business risk prediction device based on setting and adjusting a risk threshold value based on a break-even point, comprising: a memory and at least one processor, the memory stores instructions; the at least one processor calls the instructions in the memory, so that the business risk prediction device based on setting and adjusting a risk threshold value based on a break-even point executes the steps of the business risk prediction method based on setting and adjusting a risk threshold value based on a break-even point.

[0013] The fourth aspect of the present application provides a computer readable storage medium, the computer readable storage medium stores instructions, when the instructions are run on a computer, the computer executes the steps of the business risk prediction method based on setting and adjusting a risk threshold value based on a break-even point.

[0014] The fifth aspect of the present application provides a computer program product, comprising computer programs / instructions, when the computer programs / instructions are executed by a processor, the steps of the business risk prediction method based on setting and adjusting a risk threshold value based on a break-even point are realized.

[0015] In the technical scheme provided by the application, an initial prediction model is constructed, historical business data is collected and the initial prediction model is trained based on the historical business data to obtain a business risk prediction model; a break-even point of a business is calculated based on historical business data, and a risk threshold is set or adjusted based on the break-even point; a target business to be predicted is obtained, the business risk prediction model is called to perform business information prediction on the target business to obtain a business information prediction value; and the business information prediction value is compared with the set or adjusted risk threshold to determine a final business risk result. The method can dynamically adjust the risk threshold of the business risk prediction method, and improve the accuracy of credit risk control and the flexibility of decision-making. The device, equipment, medium and product provided by the application also solve the corresponding technical problems. BRIEF DESCRIPTION OF DRAWINGS

[0016] The drawings described herein are used to provide further understanding of the present application, and form a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings: Figure 1 A flowchart of a first embodiment of the business risk prediction method based on setting and adjusting of a risk threshold based on a break-even point in the embodiments of the present application; Figure 2 A flowchart of a second embodiment of the business risk prediction method based on setting and adjusting of a risk threshold based on a break-even point in the embodiments of the present application; Figure 3 An actual example graph of a corresponding curve of a business risk prediction model score and overdue non-loan repayment loan ratio data in the business risk prediction method based on setting and adjusting of a risk threshold based on a break-even point in the embodiments of the present application; Figure 4 An actual example graph of a corresponding curve of a business risk prediction model score and overdue non-loan repayment loan ratio data in the business risk prediction method based on setting and adjusting of a risk threshold based on a break-even point in the embodiments of the present application; Figure 5 An actual example graph of a loan term distribution of a 10%-13% interest rate range in the business risk prediction method based on setting and adjusting of a risk threshold based on a break-even point in the embodiments of the present application; Figure 6 An actual example graph of a corresponding curve of a business risk prediction model score and overdue non-loan repayment loan ratio data in the business risk prediction method based on setting and adjusting of a risk threshold based on a break-even point in the embodiments of the present application; Figure 7 A flowchart of a third embodiment of the business risk prediction method based on setting and adjusting of a risk threshold based on a break-even point in the embodiments of the present application; Figure 8An embodiment schematic diagram of the business risk prediction device based on the break-even point setting and the risk threshold adjustment in the embodiment of the present application; Figure 9 An embodiment schematic diagram of the business risk prediction device based on the break-even point setting and the risk threshold adjustment in the embodiment of the present application; Figure 10 An embodiment schematic diagram of the computer readable medium in the embodiment of the present application. DETAILED DESCRIPTION

[0017] Exemplary embodiments of the present application will now be described more fully hereinafter with reference to the accompanying drawings. The exemplary embodiments of the present application may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these exemplary embodiments of the present application are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the application to those skilled in the art. Like reference numerals refer to like elements throughout the specification.

[0018] In the case of conforming to the technical concept of the present application, the features, structures, characteristics or other details described in a certain specific embodiment are not excluded from being combined in one or more other embodiments in a suitable manner.

[0019] In the description of the specific embodiments, the features, structures, characteristics or other details described in the present application are to make those skilled in the art fully understand the embodiments. However, it is not excluded that one or more of the specific features, structures, characteristics or other details can be practiced without the technical solution of the present application.

[0020] The flowchart shown in the accompanying drawings is only an exemplary illustration, and is not necessarily required to include all the contents and operations / steps, nor is it necessarily required to be executed in the order described. For example, some operations / steps can be further decomposed, and some operations / steps can be combined or partially combined, so that the actual execution order can be changed according to the actual situation.

[0021] The block diagram shown in the accompanying drawings is only a functional entity, and does not necessarily correspond to a physically independent entity. That is, the functional entities can be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0022] The term "and / or" or "and / or" includes all combinations of any one or more of the associated listed items.

[0023] Please refer to Figure 1In the embodiment of the present application, the first embodiment of the business risk prediction method based on the setting and adjustment of the risk threshold value based on the break-even point comprises: S101, constructing an initial prediction model, collecting and training the initial prediction model based on historical business data to obtain a business risk prediction model; It can be understood that the execution subject of the present application can be a business risk prediction device based on the setting and adjustment of the risk threshold value based on the break-even point, and can also be a terminal or a server, which is not limited here. The server is taken as an example for illustration in the embodiment of the present application.

[0024] Before performing a specific business risk prediction, the server will first construct a business risk prediction model.

[0025] In the embodiment, the business risk is predicted mainly by predicting the risk of future business data through the mode, trend and correlation in the historical data, for example, using historical data to construct a pre-training model for specific prediction. Therefore, a statistical model, a machine learning model or the like can be used to construct an initial prediction model, collect historical business data and perform preprocessing such as cleaning and sorting, and based on the preprocessed historical business data, a training set, a verification set and a test set are formed, so as to train the initial prediction model, adjust the model parameters in the initial prediction model, and obtain a business risk prediction model.

[0026] S102, calculating the break-even point of the business based on the historical business data, and setting and adjusting the risk threshold value based on the break-even point; In the embodiment provided in the embodiment, not only the identification and control of the risk are considered, but also the balance information between the cost and the benefit of the risk control fact. Therefore, in addition to the business risk prediction model trained based on the historical business data in the embodiment, the break-even point of the business based on the historical business data is calculated, and the risk threshold value is set and adjusted based on the break-even point.

[0027] In the step, the break-even point of the business is a Break Even Point (abbreviated as BEP or BE point), also known as a zero profit point, a profit guarantee point, a profit and loss critical point, a profit and loss divergence point or a yield turning point, etc. It usually refers to the output when the total sales revenue is equal to the total cost (the intersection point of the sales revenue line and the total cost line). With the limit of the break-even point, when the sales revenue is higher than the break-even point, the enterprise makes a profit, otherwise, the enterprise incurs a loss.

[0028] In the embodiment, the annual break-even point information of the current business product can be calculated by collecting customer information contained in historical business data. Since the information represents the break-even point information of the current business product, the enterprise makes a profit when the sales revenue is higher than the break-even point, and vice versa, the enterprise incurs a loss. Therefore, the risk threshold of the business can be determined based on the annual break-even point information, so that whether the business to be predicted will incur a loss can be determined more sensitively.

[0029] For example, when a specific prediction is implemented, the business risk threshold can be dynamically adjusted based on historical break-even point (BEP), and the business risk prediction value obtained by the business risk prediction model is compared with the business risk threshold, so that the final business risk prediction result is obtained. Taking the prediction of credit risk in the financial field as an example, when the risk threshold of the business is dynamically adjusted, the risk threshold can be specifically set or adjusted based on the single loan break-even yield (hereinafter referred to as loan BE).

[0030] In the step of calculating the loan break-even point, first, the credit customer data is collected, covering basic information, default history, loan information (interest rate, amount, term, repayment method, etc.). At the same time, the latest model score is backtracked, and the FTP (Funds Transfer Pricing, internal funds transfer pricing) cost, marketing cost, data cost, and average turnover rate of each loan term are calculated. Taking a specific financial product as an example, the average customer interest rate (A) of the product is set in the interval of 10%-13%, the average value is 11.72%, the FTP cost (B) of the product is 2.5%, the marketing cost (C) is 3%, and the data cost (D) is 0.3%. Based on these cost parameters, the annual break-even yield of the product, i.e. the annual yield (E) (also referred to as loan BE), is calculated, and the formula is as follows: Annual yield (E) = average customer interest rate (A) - FTP cost (B) - marketing cost (C) - data cost (D); A specific data example is shown in Table 1: Table 1: Loan BE calculation example table Among them, the turnover rate is related to the loan term, for example, for a product with a loan term of 12 months, the average turnover rate is 2.2.

[0031] Accordingly, the single loan break-even yield (hereinafter referred to as loan BE) is further calculated, and the calculation formula of the loan BE is specifically as follows: Loan BE = annual yield (E) / average turnover rate (F); Through the above calculation, the BE of loss can be accurately evaluated under different loan term structures of the financial product, thereby providing decision support for risk management. For the product with a loan term of 12 months, the BE of loss is 2.9%. Then, a specific risk threshold can be set or adjusted according to the BE of loss.

[0032] S103, obtaining a target business to be predicted, calling a business risk prediction model to perform business information prediction on the target business, and obtaining a business information prediction value; After obtaining the business risk prediction model, target business information to be predicted can be obtained, and business information prediction can be performed on the target business information by the business risk prediction model to obtain a business information prediction value.

[0033] In a specific embodiment, the business information prediction value can be a yield rate information. When the business information prediction value is the yield rate information, the yield rate information of the current target business can be predicted by obtaining customer information, business period number, business interest rate, or repayment method information of the target business.

[0034] S104, comparing the business information prediction value with the adjusted risk threshold value, and determining a final business risk result.

[0035] After obtaining the business information prediction value, the business information prediction value is compared with the adjusted risk threshold value in the S102 step to determine a final business risk result.

[0036] For example, the annualized break-even point information is taken as the risk threshold value. When the business information prediction value of the predicted target business is higher than the risk threshold value, the risk is small, and the enterprise has a high possibility of profit. When the business information prediction value of the predicted target business is lower than the risk threshold value, there is a risk, and the enterprise has a loss risk.

[0037] In a preferred embodiment, the embodiment further includes comparing the business information prediction value with the adjusted risk threshold value, calculating the offset degree between the business information prediction value and the risk threshold value, comparing the calculated offset degree with a risk scoring rule, and generating a risk prediction result.

[0038] The embodiment of the present application proposes a business risk prediction method capable of dynamically setting a business risk threshold value. The method can consider the balance between cost and benefit, has dynamic adjustment capability, can adjust the prediction result by dynamically setting the business risk threshold value according to the actual change of the business information to be predicted, can make the result of risk assessment more accurate, improve the precision of risk control and the flexibility of decision-making, and optimize the effect of risk prediction.

[0039] Please refer to Figures 2-7The second embodiment of the business risk prediction method based on setting and adjusting risk thresholds according to the break-even point in this invention includes: S201. Construct an initial prediction model, collect and train the initial prediction model based on historical business data to obtain a business risk prediction model; It is understood that the executing entity of this invention can be a business risk prediction device that sets and adjusts risk thresholds based on a break-even point, or it can be a terminal or a server; the specific implementation is not limited here. This embodiment of the invention uses a server as an example for illustration. Specifically, this embodiment uses credit risk prediction in a financial scenario as an example for illustration.

[0040] Before the server performs specific business risk prediction, this embodiment first trains a business risk prediction model based on historical business information. This business risk model can be a KGB (Known Good Bad) model. Specifically, an initial prediction model is first constructed based on artificial intelligence algorithms. In one specific implementation, the artificial intelligence algorithm can be the Light GBM algorithm, which stands for Light Gradient Boosting Machine, a distributed gradient boosting framework based on decision tree algorithms.

[0041] Subsequently, historical business data is acquired, and training, validation, and test sets are formed based on the historical business data to train and validate the initial prediction model. The model parameters and hyperparameters included in the initial prediction model are adjusted, and finally, the business risk prediction model capable of intelligent business risk prediction tasks in this embodiment is obtained.

[0042] When implementing specific forecasts, the business risk threshold can be dynamically adjusted based on the historical break-even point (BEP). The business risk forecast value obtained from the specific forecast model is compared with the business risk threshold to obtain the final business risk forecast result. Taking credit risk forecasting in a financial scenario as an example, when dynamically adjusting the business risk threshold, the risk threshold can be specifically set or adjusted based on the break-even rate of return for a single loan (hereinafter referred to as the loss BE). The specific setting and adjustment methods are basically the same as those in step S102 of the aforementioned embodiment, so they will not be repeated here.

[0043] However, if the long-term loss-making BE is directly used as a decision indicator, the observation period will be too long. For example, for a product with a loan term of 12 months, it takes 15 months to observe the complete 90-day overdue amount overdue rate, that is, the MOB15 90day+ Amount% at the end of the 15th month after the loan month; and for a product with a loan term of 24 months, it takes 27 months to observe the complete 90-day overdue amount overdue rate, that is, the MOB27 90day+ Amount% at the end of the 27th month after the loan month. In the face of new risk trend changes, such a solution will have a serious lag in risk judgment. Therefore, in the embodiment, to reduce the difficulty of obtaining the breakeven point and obtain a more reasonable breakeven point, the second breakeven point can be obtained by the first breakeven point; wherein the second business term is shorter than the first business term, for example, the first business term is a long-term business term, and the second business term is a short-term business term.

[0044] S202, historical business data of a first business term is obtained, and a first breakeven point of the historical business data of the first business term is calculated; S203, a second business term is determined based on risk prediction requirements, and a second breakeven point of the second business term is calculated based on the first breakeven point; In the embodiment, the second breakeven point of the second business term which is relatively short-term can be inferred based on the first breakeven point calculated for the long-term first business term in the historical data, and the risk judgment threshold can be set or adjusted according to the second breakeven point, so that the prediction risk of the current business can be determined without waiting for a long observation period.

[0045] In a specific implementation, the way to calculate the second breakeven point of the shorter-term second business term can be to calculate the second breakeven point corresponding to the second business term with a length of 6 periods or 3 periods based on the first business term with a length of 12 periods or 24 periods. For example, according to a business product with a loan term of 12 months, the percentage data of the loan principal balance of the account overdue for 90 days or more in the 15-month account age to the total amount of the original loan of the batch is calculated to obtain the first breakeven point; and the second breakeven point of the shorter-term second business term is calculated according to the first breakeven point; for example, the breakeven point of the account (MOB6 30+) overdue for 30 days or more in the 6-month account age, or even the breakeven point of the account (MOB3 30+) overdue for 30 days or more in the 3-month account age.

[0046] When calculating the second break-even point, the corresponding value can be found by comparing the long-term break-even point model score with the delinquency rate percentage curve. For example, using the values ​​in Table 1 of the aforementioned embodiment, the break-even point (BE) value is determined. Please refer to the appendix. Figure 3 Taking a current loss BE (Better-Off) value of 2.9% as an example, we retrieve the model score binning data of the business risk prediction model trained above and the corresponding curve of MOB1590day+Amount% (the percentage of outstanding principal balance of loans overdue for 90 days or more within 15 months), and find the model score binning information corresponding to a loss BE value of 2.9%. (See attached...) Figure 3 The curves showing the corresponding model fraction bins (horizontal axis) and MOB15 90day+Amount% (secondary axis) reveal that the threshold for the model bin data when the loss BE value is 2.9% is 13.5.

[0047] Please continue to refer to the appendix. Figure 4 Obtain as attached Figure 4 The curve showing the model score bins (horizontal axis) and the corresponding curve of MOB6 30day+ % (the delinquency rate of accounts overdue for 30 days or more in 6 months, i.e., the content displayed on the secondary axis) allows us to find the threshold of MOB6 30+ % when the model score threshold is 13.5. In the example, the threshold is 1.29%, so we can know that the second break-even point for the second business period is 1.29%.

[0048] To illustrate with another example, for a loan with a 24-month term, we need to look at historical loan data covering 27 months. Using the same method as in the example above, we can find the second break-even point (BE) for an account with MOB3 30+% (the delinquency rate of 3 months overdue for 30 days or more). The following chart shows the second break-even point (BE) for MOB6 30+% and MOB3 30+% for various loan terms within the 10%-13% interest rate range. Please refer to Table 2; in a specific example, the BE for different loan terms and the second break-even point (BE) for shorter terms are as follows: Table 2: Break-even point information for different loan terms and shorter loan terms In a preferred embodiment, please refer to Figure 5After obtaining the second break-even point, a weighted optimization of the second break-even point is further included using the loan amount proportion in the last three months as the weight. Assuming that the future loan distribution is similar to the loan distribution in the last three months, if the future loan distribution is explicitly known, the loan plan data can be directly used; wherein, in this weighted optimization scheme, historical data is generally not used because the loan proportion may change over time. (For example, a product has a 24-period loan term starting from August 2023. If the amount before August 2023 is weighted, the break-even point of 10%-13% interest rate range obtained will not be accurate.) After determining the short-term second break-even point of each interest rate range, a reasonable risk threshold is found using the second break-even point.

[0049] S204, based on the second break-even point, setting the risk threshold corresponding to the second business term; First, according to the second break-even point of MOB6 30+, combined with the loan with 6-month performance in recent period, or using the loan with 3-month performance in recent period and MOB3 30+ second break-even point to set the risk threshold. In a specific example, the short-term break-even point information of different interest rate ranges can be as shown in Table 3: Table 3: Short-term break-even point of different interest rate ranges Assuming that the business risk prediction model's prediction of the risk of rejected samples is consistent with the risk of approved loan samples, please refer to the attached Figure 6 The risk threshold displayed on the horizontal axis corresponding to the maximum profit of MOB6 30+% BE (see Table 3, the specific value is 1.24) is 15. Generally, the business risk prediction model tends to underestimate the risk of rejected customers. Therefore, the risk of rejected customer samples placed in the same score segment can be evaluated by multiplying the risk of approved loan samples by a coefficient; in a specific way, the coefficient can be 2.

[0050] S205, obtaining the business risk preference of the target business to be predicted, and selecting a risk strategy based on the business risk preference, and adjusting the risk threshold according to the risk strategy; In this embodiment, the risk threshold can also be dynamically adjusted. As the market environment changes, the financial institution's business strategy adjusts, and the customer behavior evolves, the prediction results of the risk model and the cost-benefit analysis may change. Therefore, it is necessary to recalculate the break-even point regularly, and set or adjust the risk threshold according to the new break-even point, in order to maintain the balance between risk and return, and ensure the sustainable development of the enterprise.

[0051] In a preferred embodiment, after the risk threshold result is calculated, further threshold adjustment can be performed. Specifically, the risk tolerance of each institution can be evaluated, the business risk preference of the target business to be predicted is obtained, and a risk strategy is selected according to the business risk preference, and the risk threshold is fine-tuned according to the risk strategy. When the risk strategy is a conservative strategy, the risk threshold is lowered; when the risk strategy is a fund strategy, the risk threshold is raised.

[0052] Specifically, the risk threshold obtained can be further adjusted by weighting based on the business risk preference information obtained from the risk tolerance evaluation result. For example, for a financial institution with low risk tolerance, the threshold can be set within the break-even point to ensure strictness of risk control; for a financial institution with high risk tolerance, the threshold can be appropriately raised, and the specific threshold is determined by calculating the income and loss brought by the high-risk population, that is, more business opportunities are exchanged for slightly higher risk.

[0053] S206, obtaining a target business to be predicted, calling a business risk prediction model to predict business information of the target business, and obtaining a business information prediction value; After obtaining the business risk prediction model, the target business information to be predicted can be obtained, and the business information prediction value is obtained by predicting the business information of the target business according to the target business information through the business risk prediction model.

[0054] In a specific embodiment, the business information prediction value can be a yield rate information; when the business information prediction value is the yield rate information, the yield rate information of the current target business can be predicted by obtaining the customer information, the business period, the business interest rate or the repayment method of the target business and the like.

[0055] S207, comparing the business information prediction value and the set adjusted risk threshold to determine the final business risk result.

[0056] After obtaining the business information prediction value, the business information prediction value is compared with the risk threshold set or adjusted in the foregoing step to determine the final business risk result.

[0057] This invention proposes a business risk prediction method capable of dynamically setting business risk thresholds. This method considers the balance between costs and benefits and has dynamic adjustment capabilities. By dynamically setting business risk thresholds, the prediction results can be adjusted according to the actual changes in the business information to be predicted, making the risk assessment results more accurate, improving the precision of risk control and the flexibility of decision-making, and optimizing the effect of risk prediction. Furthermore, this embodiment also infers short-term break-even point information based on long-term break-even point information, and determines the risk threshold of business information based on the break-even point information. This can promptly capture new risk characteristics and business needs, further avoiding risk assessment deviations caused by unreasonable thresholds.

[0058] Please refer to Figure 7 The third embodiment of the business risk prediction method based on setting and adjusting risk thresholds according to the break-even point of the present invention includes: S701. Construct a credit risk control model; Before implementing specific business risk prediction and risk control steps, a credit risk control model should be built first. The model construction steps can be as described in step S201 of the aforementioned embodiments, and will not be repeated here.

[0059] S702, Input economic parameters; The economic parameters can be raw or processed parameter information contained in historical business data. These economic parameters include, but are not limited to, credit customer data, covering basic information, default history, loan information (interest rate, amount, term, repayment method, etc.). Subsequently, based on these input economic parameters, the break-even point and other information required can be calculated. For details, please refer to step S102 in the aforementioned embodiment, so it will not be repeated here.

[0060] S703, Calculate the annual return BE; This step specifically includes calculating the annual return break-even point (BE) based on historical business data as the first break-even point. The calculation method can be as described in step S102 of the aforementioned embodiment, and will not be repeated here.

[0061] S704, Derive the break-even point of MOB7 30+; In a specific example of this embodiment, the calculation of the shorter-term second break-even point based on the long-term first break-even point can be based on the MOB7 30+ break-even point derived from the annual break-even point (BE).

[0062] The derivation method is also based on the numerical lookup of the risk prediction model of the MOB7 30+ at the corresponding numerical value of the model score bin. The specific steps and explanations are basically the same as the contents in steps S202 and S203 in the foregoing embodiments, and thus will not be described here again.

[0063] S705, interest rate range weighted calculation; The embodiment also includes weighted calculation based on the interest rate range after obtaining the second break-even point. The specific weighted calculation method is basically the same as the part of the weighted calculation in the content in S203 in the foregoing embodiments, and thus will not be described here again.

[0064] S706, determining whether the loan term is clear; Determine whether the current target business has a clear business term. If yes, execute step S707, and if no, execute step S708, wherein the business term can be the loan term.

[0065] S707, setting risk threshold according to term dimension; S708, setting risk threshold according to interest rate dimension; S709, determining the final risk threshold; In the embodiment, since the specific loan business for judging risk can have multiple situations such as known interest rate but unknown specific term length and known term but unclear interest rate information, if the current target business has a clear business term, the risk threshold is set according to the term dimension; if there is no clear business term, the second break-even point is calculated based on the same information in the historical business data and the interest rate dimension, and the risk threshold is set.

[0066] S710, deploying risk control decision; Deploy the risk control decision based on the set and adjusted risk threshold, that is, calculate the possible business risk according to the actual target business information, so as to judge whether there is a business risk based on the deployed risk control decision. The specific content can refer to the contents in S205 and S206 in the foregoing embodiments, and thus will not be described here again.

[0067] S711, monitoring actual default rate and judging whether it exceeds the default rate threshold; In the embodiment, after executing the specific business risk prediction, the information of the business for which the risk prediction has been executed is monitored to obtain the actual default rate and judge whether it exceeds the default rate threshold. If yes, adjust the economic parameters in step S702 to correct the break-even point data for subsequent updating of the threshold. If no, execute step S712.

[0068] S712, periodically recalculating; Periodically reacquire the business information in a period of time to recalculate the break-even point information, and update the economic parameters in the S702 step based on the recalculated break-even point information to periodically correct the break-even point data and the risk threshold of the business risk prediction model, so that the risk prediction effect is more accurate.

[0069] The business risk prediction method provided in the embodiments of the present application can dynamically set the business risk threshold, can consider the balance of cost and benefit, has dynamic adjustment capability, can adjust the prediction result by following the actual change of the business information to be predicted through the way of dynamically setting the business risk threshold, can make the result of risk assessment more accurate, improve the accuracy of risk control and the flexibility of decision-making, and optimize the effect of risk prediction.

[0070] The business risk prediction method based on the break-even point setting and adjustment of the risk threshold in the embodiments of the present application is described above, and the business risk prediction device based on the break-even point setting and adjustment of the risk threshold in the embodiments of the present application is described below, please refer to Figure 8 The business risk prediction device based on the break-even point setting and adjustment of the risk threshold in the embodiments of the present application comprises one embodiment of: The model construction module 801 is configured to construct an initial prediction model, collect and train the initial prediction model based on historical business data to obtain a business risk prediction model; The threshold adjustment module 802 is configured to calculate the break-even point of the business based on the historical business data, and set and adjust the risk threshold based on the break-even point; The prediction execution module 803 is configured to acquire a target business to be predicted, call the business risk prediction model to perform business information prediction on the target business, and obtain a business information prediction value; The result output module 804 is configured to compare the business information prediction value with the set and adjusted risk threshold, and determine a final business risk result.

[0071] The business risk prediction method provided in the embodiments of the present application can dynamically set the business risk threshold, can consider the balance of cost and benefit, has dynamic adjustment capability, can adjust the prediction result by following the actual change of the business information to be predicted through the way of dynamically setting the business risk threshold, can make the result of risk assessment more accurate, improve the accuracy of risk control and the flexibility of decision-making, and optimize the effect of risk prediction.

[0072] In another embodiment of the present application, the threshold adjustment module 802 is specifically configured to: acquire the historical business data of the first business period, and calculate the first break-even point of the historical business data of the first business period; determine a second business duration based on the risk prediction requirement, and calculate a second break-even point of the second business duration based on the first break-even point; The second business duration has a time length less than the first business duration.

[0073] In another embodiment of the present application, the threshold adjustment module 802 is specifically further configured to: based on the historical business data, obtain a business risk prediction model risk critical point corresponding to the break-even point of the first business duration, to obtain a critical score range; based on the critical score range, find a second break-even point corresponding to the second business duration; The setting of the adjustment risk threshold based on the break-even point comprises: based on the second break-even point, setting an adjustment risk threshold corresponding to the second business duration.

[0074] In another embodiment of the present application, the first business duration is 12 periods or 24 periods, and the second business duration is 6 periods or 3 periods.

[0075] In another embodiment of the present application, the threshold adjustment module 802 is specifically further configured to: determine whether the target business content has a clear business return period; If not, obtain a historical loan amount proportion in a preset period, and perform interest rate dimension weighting on the second break-even point based on the historical loan amount proportion as a weight.

[0076] In another embodiment of the present application, the threshold adjustment module 802 is specifically further configured to: obtain a business risk preference of a target business to be predicted, and select a risk strategy based on the business risk preference; when the risk strategy is a conservative strategy, the risk threshold is lowered; when the risk strategy is a fund strategy, the risk threshold is raised.

[0077] The business risk prediction method can consider the balance of cost and benefit, has dynamic adjustment capability, can adjust the prediction result by dynamically setting the business risk threshold, can make the risk assessment result more accurate, improve the accuracy of risk control and the flexibility of decision-making, and optimize the effect of risk prediction. In the embodiment, the long-term break-even point information is used to predict the short-term break-even point information, and the risk threshold of the business information is determined based on the break-even point information, so that new risk characteristics and business requirements can be captured in time, and risk assessment deviation caused by unreasonable threshold can be further avoided.

[0078] Based on the same inventive concept, the embodiments of the present specification also provide an electronic device for business risk prediction based on setting and adjusting risk threshold value at breakeven point. The electronic device for business risk prediction based on setting and adjusting risk threshold value at breakeven point in the embodiments of the present application is described in detail from the perspective of hardware processing as follows.

[0079] Figure 9 A structural schematic diagram of an electronic device provided by the embodiments of the present specification is shown. The electronic device 900 according to the embodiments of the present application is described below with reference to Figure 9 Figure 9 The electronic device 900 shown is merely an example and should not impose any limitation on the functions and use range of the embodiments of the present application.

[0080] As shown in Figure 9 , the electronic device 900 is in the form of a general computing device. The components of the electronic device 900 can include, but are not limited to, at least one processing unit 910, at least one storage unit 920, a bus 930 connecting different system components (including the storage unit 920 and the processing unit 910), a display unit 940, etc.

[0081] The storage unit stores program codes which can be executed by the processing unit 910, so that the processing unit 910 performs the steps according to various exemplary embodiments of the present application described in the processing method part of the present specification. For example, the processing unit 910 can perform the steps as shown in Figure 1 , Figure 2 and Figure 7 .

[0082] The storage unit 920 can include a readable medium in the form of a volatile storage unit, such as a random access memory (RAM) 9201 and / or a cache memory unit 9202, and can further include a read-only memory (ROM) 9203.

[0083] The storage unit 920 can also include program / utilities 9204 with a set of (at least one) program modules 9205, such as an operating system, one or more application programs, other program modules, and program data, each of which or some combination of which can include implementation of a network environment.

[0084] The bus 930 can represent one or more of several types of bus structures, including a storage unit bus or storage unit controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of a variety of bus structures.

[0085] ​The electronic device 900 can also communicate with one or more external devices 100 such as a keyboard or pointing device, a Bluetooth device, or a device for reading media. Communication with one or more devices can enable a user to interact with the electronic device 900 in order to use the electronic device 900. In Figure 9 addition, the electronic device 900 can communicate with one or more networks of devices such as a local area network (LAN), a wide area network (WAN), or an All-IP network, for example. As depicted, the electronic device 900 can communicate with one or more networks through a network adapter 960. It should be appreciated that the network adapter 960 can be communicatively coupled to the other components of the electronic device 900 through the bus 930, although the network adapter 960 is illustrated as being a separate component in FIG. 9 to more particularly emphasize its

[0086] From the foregoing description, it will be apparent to those skilled in the art that a variety of modifications and alterations can be made to the present application without deviating from the spirit and scope of the application. Therefore, the present application should not be limited by the above described embodiments. It is understood that the applications described in this specification include all such modifications and alterations and fall within the scope of the application comprising the presently disclosed examples, together with their equivalents. Figure 1 , Figure 2 and Figure 7

[0087] Figure 10 A schematic diagram of a computer readable medium according to an embodiment of the present specification.

[0088] The present application can be implemented in other specific forms without departing from the spirit or essential characteristics thereof. Accordingly, reference should be made to the following claims, rather than to the foregoing specification as indicating the scope of the application. Figure 1 , Figure 2 and Figure 7 ​The computer program of the illustrated method can be stored on one or more computer readable media. The computer readable media can be a readable signal medium or a readable storage medium. A readable storage medium, for example, can be, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0089] The computer readable storage medium can include a data signal transported, propagated or transmitted, in baseband or as part of a carrier wave, in which the readable program code is embodied. The data signal can be transported, propagated or transmitted, for example, in a wired medium, in a wireless medium, or in any suitable combination of the foregoing. The computer readable medium can also be any medium that can be read by an instruction execution system, device or machine. The computer readable medium can include, but is not limited to, a readable storage medium or a readable signal medium.

[0090] In addition, the present application also provides a computer program product, comprising computer programs / instructions, which, when executed by a processor, implement the business risk prediction method based on the break-even point setting and risk threshold adjustment as described in any of the above embodiments.

[0091] The program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device or entirely on the remote computing device or server. In the latter scenario, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computing device, such as through the Internet using an Internet Service Provider. The present application can be implemented in a computing device that includes a processing unit, a system memory, and a system bus.

[0092] In summary, the present application can be implemented in hardware, or implemented in software modules running on one or more processors, or implemented in a combination thereof. Those skilled in the art should understand that some or all of the functions of some or all of the components according to the embodiments of the present application can be implemented in practice using a general-purpose data processing device such as a microprocessor or a digital signal processor (DSP). The present application can also be implemented as a device or apparatus program (for example, a computer program and a computer program product) for performing part or all of the methods described herein. Such a program implementing the present application can be stored on a computer readable medium or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0093] The above-described specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the present application is not inherently related to any specific computer, virtual device or electronic device, and various general-purpose devices can also implement the present application. The above-described specific embodiments are merely examples of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0094] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0095] If the technical solutions of the present application involve personal information, the product applying the technical solutions of the present application has explicitly informed the personal information processing rules before processing the personal information, and has obtained the personal independent consent. If the technical solutions of the present application involve sensitive personal information, the product applying the technical solutions of the present application has obtained the personal independent consent before processing the sensitive personal information, and at the same time meets the requirement of "explicit consent". For example, at the personal information collection device such as camera, a clear and prominent sign is set to inform that the personal information collection range has been entered, and the personal information will be collected. If the individual voluntarily enters the collection range, it is considered to agree to collect the personal information. Or, on the device for processing personal information, the personal information processing rules are informed through obvious signs / information, and the personal authorization is obtained through pop-up information or by asking the individual to upload his / her personal information. The personal information processing rules can include personal information processor, personal information processing purpose, processing method, and personal information type, etc.

[0096] The above merely provides an example of the present application, but is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application should be included in the scope of claims of the present application.

Claims

1. A business risk prediction method based on setting and adjusting risk thresholds according to the break-even point, characterized in that, include: An initial prediction model is constructed, and the initial prediction model is trained based on historical business data to obtain a business risk prediction model. Calculate the break-even point of the business based on historical business data, and set an adjustment risk threshold based on the break-even point. Obtain the target business to be predicted, call the business risk prediction model to predict the business information of the target business, and obtain the business information prediction value; The predicted business information value is compared with the adjusted risk threshold to determine the final business risk result.

2. The business risk prediction method based on setting and adjusting risk thresholds according to claim 1, characterized in that, The break-even point for the business calculated based on historical business data includes: Obtain historical business data for the first business period and calculate the first break-even point for the historical business data for the first business period; The second business term is determined based on risk prediction requirements, and the second break-even point for the second business term is calculated based on the first break-even point. The duration of the second business is shorter than that of the first business.

3. The business risk prediction method based on setting and adjusting risk thresholds according to claim 2, characterized in that, The calculation of the second break-even point for the second business period based on the first break-even point includes: Based on the historical business data, the risk threshold of the business risk prediction model corresponding to the break-even point of the first business period is obtained, and the critical score range is obtained. Based on the critical score range, find the second break-even point corresponding to the second business period; The step of setting the risk adjustment threshold based on the break-even point includes: Based on the second break-even point, a risk threshold corresponding to the second business term is set for adjustment.

4. The business risk prediction method based on setting and adjusting risk thresholds according to claim 3, characterized in that, The term of the first business is 12 or 24 periods, and the term of the second business is 6 or 3 periods.

5. The business risk prediction method based on setting and adjusting risk thresholds according to claim 3, characterized in that, After finding the second break-even point corresponding to the second business period based on the critical score range, the method further includes: Determine whether the target business content has a clearly defined business return period; If not, the historical loan amount percentage within the preset time period is obtained, and the second break-even point is weighted by the interest rate dimension based on the historical loan amount percentage.

6. The business risk prediction method based on setting and adjusting risk thresholds according to any one of claims 1-5, characterized in that, Before comparing the predicted business information value with the adjusted risk threshold to determine the final business risk result, the method further includes: Obtain the business risk appetite of the target business to be predicted, and select risk strategies based on the business risk appetite; When the risk strategy is a conservative strategy, the risk threshold is lowered; When the risk strategy is a fund strategy, the risk threshold is increased.

7. A business risk prediction device based on setting and adjusting risk thresholds according to a break-even point, characterized in that, The business risk prediction device based on setting and adjusting risk thresholds according to the break-even point includes: The model building module is used to build an initial prediction model, collect and train the initial prediction model based on historical business data, and obtain a business risk prediction model. The threshold adjustment module is used to calculate the break-even point of the business based on historical business data, and set the adjustment risk threshold based on the break-even point. The prediction execution module is used to obtain the target business to be predicted, call the business risk prediction model to predict the business information of the target business, and obtain the business information prediction value. The result output module is used to compare the predicted value of the business information with the adjusted risk threshold to determine the final business risk result.

8. A business risk prediction device based on setting and adjusting risk thresholds according to a break-even point, characterized in that, The business risk prediction device based on setting and adjusting risk thresholds according to the break-even point includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor invokes the instructions in the memory to cause the business risk prediction device based on setting and adjusting risk thresholds based on break-even point to perform the steps of the business risk prediction method based on setting and adjusting risk thresholds based on break-even point as described in any one of claims 1-7.

9. A computer-readable storage medium storing a computer program / instructions thereon, characterized in that, When the program / instruction is executed by the processor, it implements the steps of the business risk prediction method based on setting and adjusting risk thresholds according to any one of claims 1-7.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps of the business risk prediction method based on setting and adjusting risk thresholds according to any one of claims 1-7.

Images